This article is no longer actively maintained. While it remains accessible for reference, exercise caution as the information within may be outdated. Use it judiciously and consider verifying its content in light of the latest developments.
A reader has a problem with excessive leakage currents and has found nothing here. Yes, this is a concern, but what can we do?
The standards reflect little about it, and where they do, they stipulate things which either miss the point or are absolutely impossible to implement.
The most interesting standardisation working groups are those dealing with standards that do not yet exist but should exist, i. e. deal with new, previously unknown problems.
This topic also includes yet another language concern: In Germany we have founded another term for leakage current (Leckstrom) which expresses that we are not dealing with a leak but with a deliberate, “normal” operating current (Ableitstrom), which abuses the PE (protective earth) conductor that was designed to carry fault currents only. Now, if you go to www.electropedia.org (the new portal of IEV), then you find “leakage current” in either case.
There is a section in the UK wiring regulations - 543.7 - which replaces an earlier section - 607 - that covered the 'special circumstances' of earth leakage.
Basically, it states that if the leakage current for a circuit (i.e total for all appliances on the circuit) is, or may reasonably expected to be, above 10 mA, something must be done to make the PE conductor more resilient, such as dual conductors, larger sized PE plus a metal conduit or wire screen, or using an isolation transformer to localise the leakage. This applies to virtually all final circuits these days; we use ring circuits - so there are two PE conductors, and fittings with dual PE connectors (only one wire in each) - which satisfies the regulation. Then the user plugs four leaky
appliances into a multiway block with a single dodgy PE conductor - but the standards don't cover that!
Of course, we live in a culture where, for over 50 years, the PE and N are believed to be different from each other and are 'never' knowingly connected in an installation. It may be more difficult in a PEN oriented culture.
I have an interesting leakage problem. I bought two network attached storage hard drives which are in metal cases and have external plug-type power modules (with earth pins). They weigh nearly nothing, so no transformer. The power units are marked as Class 2 and CE compliant. When connected to the drive box and powered, the case sits at about 95V wrt the 'earth' potential of other devices. The short circuit current between the case and 'earth' is just under 3.5 mA. So this is a standard poor quality mains filter with the centre point connected to the dc output, but NOT to the input PE! So, is the Class 2 marking valid - does the output connection count as an exposed conducting part - and is the CE marking legal? The main point that this highlights is that the disk box itself does not need to comply with anything despite being part of a system that can expose the user to risk.
I have connected the cases to earth, of course, but these are 'high street' items from a well known distributor (but an unknown Chinese manufacturer) and there must be rather a lot of them in use.
The clause mentioned by David comes from IEC (IEC 364 Art.707.4220.127.116.11) which can also be found in Italian standards.